Bo Li

Bo Li, Haotian Liu, Liangyu Chen et al. · stanford

Advancements in large pre-trained generative models have expanded their potential as effective data generators in visual recognition. This work delves into the impact of generative images, primarily comparing paradigms that harness external data (\ie generative \vs retrieval \vs original). Our key contributions are: \textbf{1) GenBench Construction:} We devise \textbf{GenBench}, a broad benchmark comprising 22 datasets with 2548 categories, to appraise generative data across various visual recognition tasks. \textbf{2) CLER Score:} To address the insufficient correlation of existing metrics (\eg, FID, CLIP score) with downstream recognition performance, we propose \textbf{CLER}, a training-free metric indicating generative data's efficiency for recognition tasks prior to training. \textbf{3) New Baselines:} Comparisons of generative data with retrieved data from the same external pool help to elucidate the unique traits of generative data. \textbf{4) External Knowledge Injection:} By fine-tuning special token embeddings for each category via Textual Inversion, performance improves across 17 datasets, except when dealing with low-resolution reference images. Our exhaustive benchmark and analysis spotlight generative data's promise in visual recognition, while identifying key challenges for future investigation.

Jie Cheng, Yingbing Chen, Xiaodong Mei et al.

In recent years, imitation-based driving planners have reported considerable success. However, due to the absence of a standardized benchmark, the effectiveness of various designs remains unclear. The newly released nuPlan addresses this issue by offering a large-scale real-world dataset and a standardized closed-loop benchmark for equitable comparisons. Utilizing this platform, we conduct a comprehensive study on two fundamental yet underexplored aspects of imitation-based planners: the essential features for ego planning and the effective data augmentation techniques to reduce compounding errors. Furthermore, we highlight an imitation gap that has been overlooked by current learning systems. Finally, integrating our findings, we propose a strong baseline model-PlanTF. Our results demonstrate that a well-designed, purely imitation-based planner can achieve highly competitive performance compared to state-of-the-art methods involving hand-crafted rules and exhibit superior generalization capabilities in long-tail cases. Our models and benchmarks are publicly available. Project website https://jchengai.github.io/planTF.

Yurui Zhu, Xueyang Fu, Peng-Tao Jiang et al.

This research focuses on the issue of single-image reflection removal (SIRR) in real-world conditions, examining it from two angles: the collection pipeline of real reflection pairs and the perception of real reflection locations. We devise an advanced reflection collection pipeline that is highly adaptable to a wide range of real-world reflection scenarios and incurs reduced costs in collecting large-scale aligned reflection pairs. In the process, we develop a large-scale, high-quality reflection dataset named Reflection Removal in the Wild (RRW). RRW contains over 14,950 high-resolution real-world reflection pairs, a dataset forty-five times larger than its predecessors. Regarding perception of reflection locations, we identify that numerous virtual reflection objects visible in reflection images are not present in the corresponding ground-truth images. This observation, drawn from the aligned pairs, leads us to conceive the Maximum Reflection Filter (MaxRF). The MaxRF could accurately and explicitly characterize reflection locations from pairs of images. Building upon this, we design a reflection location-aware cascaded framework, specifically tailored for SIRR. Powered by these innovative techniques, our solution achieves superior performance than current leading methods across multiple real-world benchmarks. Codes and datasets will be publicly available.

Zhiyang Jia, Hongyan Cui, Ge Gao et al.

Person re-identification (ReID) plays a pivotal role in computer vision, particularly in surveillance and security applications within IoT-enabled smart environments. This study introduces the Enhanced Pedestrian Alignment Network (EPAN), tailored for robust ReID across diverse IoT surveillance conditions. EPAN employs a dual-branch architecture to mitigate the impact of perspective and environmental changes, extracting alignment information under varying scales and viewpoints. Here, we demonstrate EPAN's strong feature extraction capabilities, achieving outstanding performance on the Inspection-Personnel dataset with a Rank-1 accuracy of 90.09% and a mean Average Precision (mAP) of 78.82%. This highlights EPAN's potential for real-world IoT applications, enabling effective and reliable person ReID across diverse cameras in surveillance and security systems. The code and data are available at: https://github.com/ggboy2580/EPAN

Tianyu Zhang, Lingxi Xie, Longhui Wei et al.

The main difficulty of person re-identification (ReID) lies in collecting annotated data and transferring the model across different domains. This paper presents UnrealPerson, a novel pipeline that makes full use of unreal image data to decrease the costs in both the training and deployment stages. Its fundamental part is a system that can generate synthesized images of high-quality and from controllable distributions. Instance-level annotation goes with the synthesized data and is almost free. We point out some details in image synthesis that largely impact the data quality. With 3,000 IDs and 120,000 instances, our method achieves a 38.5% rank-1 accuracy when being directly transferred to MSMT17. It almost doubles the former record using synthesized data and even surpasses previous direct transfer records using real data. This offers a good basis for unsupervised domain adaption, where our pre-trained model is easily plugged into the state-of-the-art algorithms towards higher accuracy. In addition, the data distribution can be flexibly adjusted to fit some corner ReID scenarios, which widens the application of our pipeline. We will publish our data synthesis toolkit and synthesized data in https://github.com/FlyHighest/UnrealPerson.

Chaoyou Fu, Yi-Fan Zhang, Shukang Yin et al. · pku

As a prominent direction of Artificial General Intelligence (AGI), Multimodal Large Language Models (MLLMs) have garnered increased attention from both industry and academia. Building upon pre-trained LLMs, this family of models further develops multimodal perception and reasoning capabilities that are impressive, such as writing code given a flow chart or creating stories based on an image. In the development process, evaluation is critical since it provides intuitive feedback and guidance on improving models. Distinct from the traditional train-eval-test paradigm that only favors a single task like image classification, the versatility of MLLMs has spurred the rise of various new benchmarks and evaluation methods. In this paper, we aim to present a comprehensive survey of MLLM evaluation, discussing four key aspects: 1) the summarised benchmarks types divided by the evaluation capabilities, including foundation capabilities, model self-analysis, and extented applications; 2) the typical process of benchmark counstruction, consisting of data collection, annotation, and precautions; 3) the systematic evaluation manner composed of judge, metric, and toolkit; 4) the outlook for the next benchmark. This work aims to offer researchers an easy grasp of how to effectively evaluate MLLMs according to different needs and to inspire better evaluation methods, thereby driving the progress of MLLM research.

Xi Wang, Xueyang Fu, Peng-Tao Jiang et al.

Adverse weather image restoration strives to recover clear images from those affected by various weather types, such as rain, haze, and snow. Each weather type calls for a tailored degradation removal approach due to its unique impact on images. Conversely, content reconstruction can employ a uniform approach, as the underlying image content remains consistent. Although previous techniques can handle multiple weather types within a single network, they neglect the crucial distinction between these two processes, limiting the quality of restored images. This work introduces a novel adverse weather image restoration method, called DDCNet, which decouples the degradation removal and content reconstruction process at the feature level based on their channel statistics. Specifically, we exploit the unique advantages of the Fourier transform in both these two processes: (1) the degradation information is mainly located in the amplitude component of the Fourier domain, and (2) the Fourier domain contains global information. The former facilitates channel-dependent degradation removal operation, allowing the network to tailor responses to various adverse weather types; the latter, by integrating Fourier's global properties into channel-independent content features, enhances network capacity for consistent global content reconstruction. We further augment the degradation removal process with a degradation mapping loss function. Extensive experiments demonstrate our method achieves state-of-the-art performance in multiple adverse weather removal benchmarks.

Bo Li, Chengben Xu, Wufeng Zhang

This paper presents Seewo's systems for both tracks of the Multilingual Conversational Speech Language Model Challenge (MLC-SLM), addressing automatic speech recognition (ASR) and speaker diarization with ASR (SD-ASR). We introduce a multi-stage training pipeline that explicitly enhances reasoning and self-correction in speech language models for ASR. Our approach combines curriculum learning for progressive capability acquisition, Chain-of-Thought data augmentation to foster intermediate reflection, and Reinforcement Learning with Verifiable Rewards (RLVR) to further refine self-correction through reward-driven optimization. This approach achieves substantial improvements over the official challenge baselines. On the evaluation set, our best system attains a WER/CER of 11.57% for Track 1 and a tcpWER/tcpCER of 17.67% for Track 2. Comprehensive ablation studies demonstrate the effectiveness of each component under challenge constraints.

Bo Li, Wei Wang, Peng Ye

Differential privacy (DP) is a formal notion that restricts the privacy leakage of an algorithm when running on sensitive data, in which privacy-utility trade-off is one of the central problems in private data analysis. In this work, we investigate the fundamental limits of differential privacy in online learning algorithms and present evidence that separates three types of constraints: no DP, pure DP, and approximate DP. We first describe a hypothesis class that is online learnable under approximate DP but not online learnable under pure DP under the adaptive adversarial setting. This indicates that approximate DP must be adopted when dealing with adaptive adversaries. We then prove that any private online learner must make an infinite number of mistakes for almost all hypothesis classes. This essentially generalizes previous results and shows a strong separation between private and non-private settings since a finite mistake bound is always attainable (as long as the class is online learnable) when there is no privacy requirement.

Jinghao Zhou, Bo Li, Lei Qiao et al.

Visual Object Tracking (VOT) has synchronous needs for both robustness and accuracy. While most existing works fail to operate simultaneously on both, we investigate in this work the problem of conflicting performance between accuracy and robustness. We first conduct a systematic comparison among existing methods and analyze their restrictions in terms of accuracy and robustness. Specifically, 4 formulations-offline classification (OFC), offline regression (OFR), online classification (ONC), and online regression (ONR)-are considered, categorized by the existence of online update and the types of supervision signal. To account for the problem, we resort to the idea of ensemble and propose a dual-modal framework for target localization, consisting of robust localization suppressing distractors via ONR and the accurate localization attending to the target center precisely via OFC. To yield a final representation (i.e, bounding box), we propose a simple but effective score voting strategy to involve adjacent predictions such that the final representation does not commit to a single location. Operating beyond the real-time demand, our proposed method is further validated on 8 datasets-VOT2018, VOT2019, OTB2015, NFS, UAV123, LaSOT, TrackingNet, and GOT-10k, achieving state-of-the-art performance.

Shaofei Huang, Si Liu, Tianrui Hui et al.

Learning to capture dependencies between spatial positions is essential to many visual tasks, especially the dense labeling problems like scene parsing. Existing methods can effectively capture long-range dependencies with self-attention mechanism while short ones by local convolution. However, there is still much gap between long-range and short-range dependencies, which largely reduces the models' flexibility in application to diverse spatial scales and relationships in complicated natural scene images. To fill such a gap, we develop a Middle-Range (MR) branch to capture middle-range dependencies by restricting self-attention into local patches. Also, we observe that the spatial regions which have large correlations with others can be emphasized to exploit long-range dependencies more accurately, and thus propose a Reweighed Long-Range (RLR) branch. Based on the proposed MR and RLR branches, we build an Omni-Range Dependencies Network (ORDNet) which can effectively capture short-, middle- and long-range dependencies. Our ORDNet is able to extract more comprehensive context information and well adapt to complex spatial variance in scene images. Extensive experiments show that our proposed ORDNet outperforms previous state-of-the-art methods on three scene parsing benchmarks including PASCAL Context, COCO Stuff and ADE20K, demonstrating the superiority of capturing omni-range dependencies in deep models for scene parsing task.

Wenting Tang, Xingxing Wei, Bo Li

In the traditional deep compression framework, iteratively performing network pruning and quantization can reduce the model size and computation cost to meet the deployment requirements. However, such a step-wise application of pruning and quantization may lead to suboptimal solutions and unnecessary time consumption. In this paper, we tackle this issue by integrating network pruning and quantization as a unified joint compression problem and then use AutoML to automatically solve it. We find the pruning process can be regarded as the channel-wise quantization with 0 bit. Thus, the separate two-step pruning and quantization can be simplified as the one-step quantization with mixed precision. This unification not only simplifies the compression pipeline but also avoids the compression divergence. To implement this idea, we propose the automated model compression by jointly applied pruning and quantization (AJPQ). AJPQ is designed with a hierarchical architecture: the layer controller controls the layer sparsity, and the channel controller decides the bit-width for each kernel. Following the same importance criterion, the layer controller and the channel controller collaboratively decide the compression strategy. With the help of reinforcement learning, our one-step compression is automatically achieved. Compared with the state-of-the-art automated compression methods, our method obtains a better accuracy while reducing the storage considerably. For fixed precision quantization, AJPQ can reduce more than five times model size and two times computation with a slight performance increase for Skynet in remote sensing object detection. When mixed-precision is allowed, AJPQ can reduce five times model size with only 1.06% top-5 accuracy decline for MobileNet in the classification task.

Shaofei Huang, Tianrui Hui, Si Liu et al.

Referring image segmentation aims at segmenting the foreground masks of the entities that can well match the description given in the natural language expression. Previous approaches tackle this problem using implicit feature interaction and fusion between visual and linguistic modalities, but usually fail to explore informative words of the expression to well align features from the two modalities for accurately identifying the referred entity. In this paper, we propose a Cross-Modal Progressive Comprehension (CMPC) module and a Text-Guided Feature Exchange (TGFE) module to effectively address the challenging task. Concretely, the CMPC module first employs entity and attribute words to perceive all the related entities that might be considered by the expression. Then, the relational words are adopted to highlight the correct entity as well as suppress other irrelevant ones by multimodal graph reasoning. In addition to the CMPC module, we further leverage a simple yet effective TGFE module to integrate the reasoned multimodal features from different levels with the guidance of textual information. In this way, features from multi-levels could communicate with each other and be refined based on the textual context. We conduct extensive experiments on four popular referring segmentation benchmarks and achieve new state-of-the-art performances.

Bo Li, Marius de Groot, Rebecca M. E. Steketee et al.

Subtle changes in white matter (WM) microstructure have been associated with normal aging and neurodegeneration. To study these associations in more detail, it is highly important that the WM tracts can be accurately and reproducibly characterized from brain diffusion MRI. In addition, to enable analysis of WM tracts in large datasets and in clinical practice it is essential to have methodology that is fast and easy to apply. This work therefore presents a new approach for WM tract segmentation: Neuro4Neuro, that is capable of direct extraction of WM tracts from diffusion tensor images using convolutional neural network (CNN). This 3D end-to-end method is trained to segment 25 WM tracts in aging individuals from a large population-based study (N=9752, 1.5T MRI). The proposed method showed good segmentation performance and high reproducibility, i.e., a high spatial agreement (Cohen's kappa, k = 0.72 ~ 0.83) and a low scan-rescan error in tract-specific diffusion measures (e.g., fractional anisotropy: error = 1% ~ 5%). The reproducibility of the proposed method was higher than that of a tractography-based segmentation algorithm, while being orders of magnitude faster (0.5s to segment one tract). In addition, we showed that the method successfully generalizes to diffusion scans from an external dementia dataset (N=58, 3T MRI). In two proof-of-principle experiments, we associated WM microstructure obtained using the proposed method with age in a normal elderly population, and with disease subtypes in a dementia cohort. In concordance with the literature, results showed a widespread reduction of microstructural organization with aging and substantial group-wise microstructure differences between dementia subtypes. In conclusion, we presented a highly reproducible and fast method for WM tract segmentation that has the potential of being used in large-scale studies and clinical practice.

Tianyu Zhang, Lingxi Xie, Longhui Wei et al.

Person re-identification (ReID) aims at finding the same person in different cameras. Training such systems usually requires a large amount of cross-camera pedestrians to be annotated from surveillance videos, which is labor-consuming especially when the number of cameras is large. Differently, this paper investigates ReID in an unexplored single-camera-training (SCT) setting, where each person in the training set appears in only one camera. To the best of our knowledge, this setting was never studied before. SCT enjoys the advantage of low-cost data collection and annotation, and thus eases ReID systems to be trained in a brand new environment. However, it raises major challenges due to the lack of cross-camera person occurrences, which conventional approaches heavily rely on to extract discriminative features. The key to dealing with the challenges in the SCT setting lies in designing an effective mechanism to complement cross-camera annotation. We start with a regular deep network for feature extraction, upon which we propose a novel loss function named multi-camera negative loss (MCNL). This is a metric learning loss motivated by probability, suggesting that in a multi-camera system, one image is more likely to be closer to the most similar negative sample in other cameras than to the most similar negative sample in the same camera. In experiments, MCNL significantly boosts ReID accuracy in the SCT setting, which paves the way of fast deployment of ReID systems with good performance on new target scenes.

Yue Liao, Si Liu, Guanbin Li et al.

Referring expression comprehension aims to localize the object instance described by a natural language expression. Current referring expression methods have achieved good performance. However, none of them is able to achieve real-time inference without accuracy drop. The reason for the relatively slow inference speed is that these methods artificially split the referring expression comprehension into two sequential stages including proposal generation and proposal ranking. It does not exactly conform to the habit of human cognition. To this end, we propose a novel Realtime Cross-modality Correlation Filtering method (RCCF). RCCF reformulates the referring expression comprehension as a correlation filtering process. The expression is first mapped from the language domain to the visual domain and then treated as a template (kernel) to perform correlation filtering on the image feature map. The peak value in the correlation heatmap indicates the center points of the target box. In addition, RCCF also regresses a 2-D object size and 2-D offset. The center point coordinates, object size and center point offset together to form the target bounding box. Our method runs at 40 FPS while achieving leading performance in RefClef, RefCOCO, RefCOCO+ and RefCOCOg benchmarks. In the challenging RefClef dataset, our methods almost double the state-of-the-art performance (34.70% increased to 63.79%). We hope this work can arouse more attention and studies to the new cross-modality correlation filtering framework as well as the one-stage framework for referring expression comprehension.

Yuchao Dai, Zhidong Zhu, Zhibo Rao et al.

The success of existing deep-learning based multi-view stereo (MVS) approaches greatly depends on the availability of large-scale supervision in the form of dense depth maps. Such supervision, while not always possible, tends to hinder the generalization ability of the learned models in never-seen-before scenarios. In this paper, we propose the first unsupervised learning based MVS network, which learns the multi-view depth maps from the input multi-view images and does not need ground-truth 3D training data. Our network is symmetric in predicting depth maps for all views simultaneously, where we enforce cross-view consistency of multi-view depth maps during both training and testing stages. Thus, the learned multi-view depth maps naturally comply with the underlying 3D scene geometry. Besides, our network also learns the multi-view occlusion maps, which further improves the robustness of our network in handling real-world occlusions. Experimental results on multiple benchmarking datasets demonstrate the effectiveness of our network and the excellent generalization ability.

Guanjun Guo, Hanzi Wang, Yan Yan et al.

Current face or object detection methods via convolutional neural network (such as OverFeat, R-CNN and DenseNet) explicitly extract multi-scale features based on an image pyramid. However, such a strategy increases the computational burden for face detection. In this paper, we propose a fast face detection method based on discriminative complete features (DCFs) extracted by an elaborately designed convolutional neural network, where face detection is directly performed on the complete feature maps. DCFs have shown the ability of scale invariance, which is beneficial for face detection with high speed and promising performance. Therefore, extracting multi-scale features on an image pyramid employed in the conventional methods is not required in the proposed method, which can greatly improve its efficiency for face detection. Experimental results on several popular face detection datasets show the efficiency and the effectiveness of the proposed method for face detection.

Xiaowei Zhang, Li Cheng, Bo Li et al.

A major bottleneck of pedestrian detection lies on the sharp performance deterioration in the presence of small-size pedestrians that are relatively far from the camera. Motivated by the observation that pedestrians of disparate spatial scales exhibit distinct visual appearances, we propose in this paper an active pedestrian detector that explicitly operates over multiple-layer neuronal representations of the input still image. More specifically, convolutional neural nets such as ResNet and faster R-CNNs are exploited to provide a rich and discriminative hierarchy of feature representations as well as initial pedestrian proposals. Here each pedestrian observation of distinct size could be best characterized in terms of the ResNet feature representation at a certain layer of the hierarchy; Meanwhile, initial pedestrian proposals are attained by faster R-CNNs techniques, i.e. region proposal network and follow-up region of interesting pooling layer employed right after the specific ResNet convolutional layer of interest, to produce joint predictions on the bounding-box proposals' locations and categories (i.e. pedestrian or not). This is engaged as input to our active detector where for each initial pedestrian proposal, a sequence of coordinate transformation actions is carried out to determine its proper x-y 2D location and layer of feature representation, or eventually terminated as being background. Empirically our approach is demonstrated to produce overall lower detection errors on widely-used benchmarks, and it works particularly well with far-scale pedestrians. For example, compared with 60.51% log-average miss rate of the state-of-the-art MS-CNN for far-scale pedestrians (those below 80 pixels in bounding-box height) of the Caltech benchmark, the miss rate of our approach is 41.85%, with a notable reduction of 18.68%.

Bo Li, Yuchao Dai, Mingyi He

Monocular depth estimation is a challenging task in complex compositions depicting multiple objects of diverse scales. Albeit the recent great progress thanks to the deep convolutional neural networks (CNNs), the state-of-the-art monocular depth estimation methods still fall short to handle such real-world challenging scenarios. In this paper, we propose a deep end-to-end learning framework to tackle these challenges, which learns the direct mapping from a color image to the corresponding depth map. First, we represent monocular depth estimation as a multi-category dense labeling task by contrast to the regression based formulation. In this way, we could build upon the recent progress in dense labeling such as semantic segmentation. Second, we fuse different side-outputs from our front-end dilated convolutional neural network in a hierarchical way to exploit the multi-scale depth cues for depth estimation, which is critical to achieve scale-aware depth estimation. Third, we propose to utilize soft-weighted-sum inference instead of the hard-max inference, transforming the discretized depth score to continuous depth value. Thus, we reduce the influence of quantization error and improve the robustness of our method. Extensive experiments on the NYU Depth V2 and KITTI datasets show the superiority of our method compared with current state-of-the-art methods. Furthermore, experiments on the NYU V2 dataset reveal that our model is able to learn the probability distribution of depth.

Hai-Miao Hu, Bo Li, Weiyao Lin et al.

Rate-control plays an important role in video coding. However, in the conventional rate-control algorithms, the number and position of Macroblocks (MBs) inside one basic unit for rate-control is inflexible and predetermined. The different characteristics of the MBs are not fully considered. Also, there is no overall optimization of the coding of basic units. This paper proposes a new region-based rate-control scheme for H.264/AVC to improve the coding efficiency. The inter-frame information is explored to objectively divide one frame into multiple regions based on their rate-distortion behaviors. The MBs with the similar characteristics are classified into the same region, and the entire region instead of a single MB or a group of contiguous MBs is treated as a basic unit for rate-control. A linear rate-quantization stepsize model and a linear distortion-quantization stepsize model are proposed to accurately describe the rate-distortion characteristics for the region-based basic units. Moreover, based on the above linear models, an overall optimization model is proposed to obtain suitable Quantization Parameters (QPs) for the region-based basic units. Experimental results demonstrate that the proposed region-based rate-control approach can achieve both better subjective and objective quality by performing the rate-control adaptively with the content, compared to the conventional rate-control approaches.